Abstract
Abstract 686
DLBCL has two molecular subtypes, termed activated B cell-like (ABC) and germinal center B cell-like (GCB), with ABC DLBCL being less curable with current therapy. The survival of ABC but not GCB DLBCL cell lines is sustained by “chronic active” B cell receptor (BCR) signaling. Gain-of-function mutations affecting the BCR subunit CD79B occur in 21% of ABC but only 5% of GCB DLBCL tumors. ABC DLBCL cell lines also depend upon a second pathway for survival that is coordinated by MYD88, an adapter for Toll-like receptors. A constitutively active MYD88 mutant (L265P) is frequent in ABC DLBCL tumors (29%) but rare in GCB DLBCL. CD79B and MYD88 L265P mutations often coexist in ABC DLBCL tumors, suggesting oncogenic collaboration, but can also occur alone. Here, we present interim results of a phase 2 study in relapsed/refractory DLBCL of ibrutinib, a first in class inhibitor of BTK, a kinase in the BCR pathway. We tested the hypothesis that ibrutinib would be more active in ABC than GCB DLBCL due to their different addiction pathways, and we assessed the association of CD79B and MYD88 mutations with response.
Subjects with relapsed/refractory de novo DLBCL received ibrutinib 560 mg PO QD. Gene expression profiling (GEP) of formalin-fixed paraffin-embedded biopsy tissues using Affymetrix arrays was used to identify the DLBCL subtype (ABC, GCB, unclassifiable) or were not arrayed (unknown). Sanger sequencing was used to identify CD79B and MYD88 mutations. Subjects underwent CT and PET scanning pre-treatment and every 2 cycles. The primary objective of the study was overall response rate (ORR) categorized by molecular subtype. Response was investigator determined using the revised International Working Group Criteria for NHL.
Seventy subjects were enrolled; median age 63 yrs (28–92); male 71%; stage IV 63%; HI-I/HI IPI 59%; disease ≥10 cm 23%; median prior systemic therapies 3 (1–7); relapsed (27%), refractory (54%) and unknown (19%); prior stem cell transplant 23%; and median time from diagnosis 19 months.
Safety data are available for 68 subjects who received ≥ 1 dose of ibrutinib. Ibrutinib was well tolerated, with treatment-emergent AEs consistent with data reported in other ibrutinib studies. No new safety signals were identified. Sixty subjects were evaluable per protocol for response (≥ 1 dose of ibrutinib and at least one response assessment). Four subjects in the ABC cohort were not evaluable for response (1 death at study day 12, 1 PD at study day 65, and 2 remain on study treatment but have not had their first response assessment at this analysis). One subject in the GCB cohort was not evaluable for response (death at study day 41). In the ABC subtype, per protocol ORR was 40% (10/25, 95% CI: 21–61%), CR 8% (2/25) and PR 32% (8/25). The median PFS at the time of this analysis is 5.5 months in ABC responders with 60% having not progressed (5 remain on treatment and 1 responder proceeded to transplant). Only one PR was observed in the GCB subtype and none in unclassifiable cases. Thus, ibrutinib showed preferential response activity in ABC versus GCB DLBCL (p=0.0126, Fisher's exact test). Responses occurred in ABC DLBCL tumors with CD79B mutations (60%; 3/5), but also in those with wild type CD79B (37%; 7/19), suggesting that ibrutinib sensitivity does not require a BCR mutation. All cases with both CD79B and MYD88 L265P mutations (n=4) responded, showing that the MYD88 pathway does not prevent ibrutinib activity. In comparison, tumors with only a MYD88 L265P mutation (n=4) did not respond (p=0.0286, Fisher's exact test), suggesting a MYD88-dependent but BCR-independent pathogenesis for some ABC DLBCL cases.
. | ABC subtype (N=29) . | GCB subtype (N=20) . | Unclassifiable1 (N=16) . | Unknown2 (N=5) . | Total (N=70) . |
---|---|---|---|---|---|
Not Evaluable for Response | 4 | 1 | 3 | 2 | 10 |
PP ORR4 (CR + PR) | 10 (40%) | 1 (5.3%) | 0 | 2 (66.7%) | 13 (21.7%) |
Complete Response (CR) | 2 (8%) | 0 | 0 | 1 (33.3%) | 3 (5%) |
Partial Response (PR) | 8 (32%) | 1 (5.3%) | 0 | 1 (33.3%) | 10 (16.7%) |
PFS (months) | 2.5 | 1.28 | 0.95 | NR3 | 1.64 |
. | ABC subtype (N=29) . | GCB subtype (N=20) . | Unclassifiable1 (N=16) . | Unknown2 (N=5) . | Total (N=70) . |
---|---|---|---|---|---|
Not Evaluable for Response | 4 | 1 | 3 | 2 | 10 |
PP ORR4 (CR + PR) | 10 (40%) | 1 (5.3%) | 0 | 2 (66.7%) | 13 (21.7%) |
Complete Response (CR) | 2 (8%) | 0 | 0 | 1 (33.3%) | 3 (5%) |
Partial Response (PR) | 8 (32%) | 1 (5.3%) | 0 | 1 (33.3%) | 10 (16.7%) |
PFS (months) | 2.5 | 1.28 | 0.95 | NR3 | 1.64 |
GEP performed, but not assignable to ABC or GCB subtypes.
GEP not yet performed or tissue not available.
Not reached.
PP = per protocol.
Ibrutinib showed a clinically meaningful response rate in relapsed/refractory ABC DLBCL, but not in other molecular subtypes. These results are consistent with an essential role of BCR signaling in ABC DLBCL and indicate that future clinical trials of ibrutinib in DLBCL should enroll patients with this subtype.
Goy:Pharmacyclics: Research Funding. de Vos:Pharmacyclics: Research Funding. Kenkre:Pharmacyclics: Research Funding. Blum:Pharmacyclics: Research Funding. Advani:Phramacyclics, Inc: Research Funding. Kunkel:Pharmacyclics, Inc: Employment, Equity Ownership. McGreivy:Pharmacyclics, Inc.: Employment, Equity Ownership. Balasubramanian:Pharmacyclics, Inc.: Employment, Equity Ownership. Cheng:Pharmacyclics, Inc.: Employment, Equity Ownership. Moussa:Pharmacyclics, Inc.: Employment, Equity Ownership. Buggy:Pharmacyclics, Inc.: Employment, Equity Ownership.
Author notes
Asterisk with author names denotes non-ASH members.